MY-012985 A long semaphore wait

[acerola]

After I upgraded from Percona 8.0.16-7 to 8.0.18-9, I started getting these warnings in the log. Also, performance is very poor. I run Centos 7. Any ideas?

2019-12-19T12:20:15.137744Z 0 [Warning] [MY-012985] [InnoDB] A long semaphore wait:
--Thread 139710936921856 has waited at dict0stats.cc line 1260 for 241 seconds the semaphore:
S-lock on RW-latch at 0x7efebf166da0 created in file buf0buf.cc line 785
a writer (thread id 139709589550848) has reserved it in mode exclusive
number of readers 0, waiters flag 1, lock_word: 0
Last time read locked in file row0sel.cc line 4922
Last time write locked in file /mnt/workspace/percona-server-8.0-redhat-binary/label_exp/min-centos-7-x64/test/rpmbuild/BUILD/percona-server-8.0.18-9/percona-server-8.0.18-9/storage/innobase/row/row0upd.cc line 3047
2019-12-19T12:20:15.146845Z 0 [Warning] [MY-012985] [InnoDB] A long semaphore wait:
--Thread 139690600232704 has waited at row0sel.cc line 4922 for 241 seconds the semaphore:
S-lock on RW-latch at 0x7efebf166da0 created in file buf0buf.cc line 785
a writer (thread id 139709589550848) has reserved it in mode exclusive
number of readers 0, waiters flag 1, lock_word: 0
Last time read locked in file row0sel.cc line 4922
Last time write locked in file /mnt/workspace/percona-server-8.0-redhat-binary/label_exp/min-centos-7-x64/test/rpmbuild/BUILD/percona-server-8.0.18-9/percona-server-8.0.18-9/storage/innobase/row/row0upd.cc line 3047
InnoDB: ###### Starts InnoDB Monitor for 30 secs to print diagnostic info:
InnoDB: Pending preads 0, pwrites 0

=====================================
2019-12-19 09:20:20 0x7f11ece1a700 INNODB MONITOR OUTPUT
=====================================

[George_Lorch]

We have not seen such behavior in our testing so far, but it is entirely possible that we are missing some specific case. There is always the possibility of a regression introduced either by us here at Percona or inherited from Oracle MySQL. Please report a bug at https://jira.percona.com/projects/PS with as much information that you can regarding your environment, schemas, data, queries, etc… and our support team will try to confirm the issue.

–
George O. Lorch III
Director of Server Engineering - Percona Server

[acerola]

I created the bug report: https://jira.percona.com/browse/PS-6756

I noticed this in the log:

2019-12-19T08:12:25.030516Z 0 [System] [MY-010910] [Server] /usr/sbin/mysqld: Shutdown complete (mysqld 8.0.16-7) Percona Server (GPL), Release 7, Revision 613e312.
2019-12-19T08:12:27.503856Z 0 [System] [MY-010116] [Server] /usr/sbin/mysqld (mysqld 8.0.18-9) starting as process 175120
2019-12-19T08:12:38.587506Z 1 [System] [MY-011090] [Server] Data dictionary upgrading from version '80016' to '80017'.
2019-12-19T08:12:39.402430Z 1 [System] [MY-013413] [Server] Data dictionary upgrade from version '80016' to '80017' completed.
2019-12-19T08:12:44.208245Z 4 [System] [MY-013381] [Server] Server upgrade from '80016' to '80018' started.
2019-12-19T08:12:49.309843Z 4 [System] [MY-013381] [Server] Server upgrade from '80016' to '80018' completed.

The data dictionary upgraded from 80016 to 80017, and not to 80018. Is this normal?